The findings confirmed the hypotheses that female romantic partners' antisocial behavior and men's deviant peer association would be linked to men's involvement in crime in early adulthood. Also as predicted, relationship stability was the key social-control factor related to desistance from persistent criminal behavior. A key contribution of the study is that factors related to social learning and social control operate in romantic relationships in general rather than only in marital relationships. Substance use and depressive symptoms showed little prediction of adult arrests in the multivariate models, although in the final model that predicted arrest count for persistence, women's depressive symptoms were predictive of the men's persistence in crime. Because of the possible damaging effects of social influence in romantic relationships on men's criminal behavior, more attention should be given to dynamics with romantic relationships in early adulthood, with attention to prevention and treatment intervention that may counter the negative effects of these relationships. Participants for the current study were from the Oregon Youth Study (OYS), a community based sample of 206 young men at risk for delinquency. The focus was on the Couples Study, which addressed the OYS men's romantic relationships. The Couples Study was a companion longitudinal study in which the OYS men and their partners were assessed five times, at ages 17-20, ages 20-23, ages 23-25, ages 25-27, and ages 27-29. The dependent variable was men's arrests in early adulthood. Time-varying covariates were men's deviant peer association, men's substance use, men's depressive symptoms, women's antisocial behavior, women's substance use, women's depressive symptoms, men's attachment to partner, and relationship stability. 3 tables and 66 references